Arc Flash Forum
https://brainfiller.com/arcflashforum/

How accurate is arc calculation really
https://brainfiller.com/arcflashforum/viewtopic.php?f=2&t=5664		 Page 1 of 1
Author:  Jensaugust12 [ Tue Feb 07, 2023 7:56 am ]
Post subject:  How accurate is arc calculation really
Are there any known studies that have looked at how accurate the calculation of incident energy really is?
(IEEE1584, Arcpro)
Author:  Jensaugust12 [ Wed Feb 08, 2023 1:46 am ]
Post subject:  Re: How accurate is arc calculation really
Anyone know if this has been looked into?
Author:  bbaumer [ Wed Feb 08, 2023 10:31 am ]
Post subject:  Re: How accurate is arc calculation really
Jim is the only one I know of on this forum that can respond with actual test results vs calculations, and the calcs were derived in part from test results.

I can say this, I have done some forensic / expert witness work on cases involving arcing faults. A few cases captured on surveillance video. I modeled one of those events that the calcs said the upstream current limiting fuses should have blown quickly but the event sustained long enough for the employees to see what was happening, approach the panel ( a 200A, 480V machine control panel ), decide that was a bad idea, and run to open the upstream disconnect. I don't remember now exactly how much time that took but it was several seconds. 10? 20? More? The video showed a sustained glow the entire time until the employee shut it down.

If Jim reads this I wonder if he can comment on unexpected test results, several standard deviations outside of the norm or tests when the OCP device failed to open when the calcs would have said otherwise.
Author:  Jensaugust12 [ Thu Feb 09, 2023 11:14 pm ]
Post subject:  Re: How accurate is arc calculation really
Thanks!
I guess it is also better to wear PPE than not..
But it would have been very interesting to see an approach to this.
How much do the various factors really matter in relation to reality.
Author:  Jim Phillips (brainfiller) [ Mon Feb 13, 2023 8:13 am ]
Post subject:  Re: How accurate is arc calculation really
I did not sift though all of the data but recall one of the many objectives for the 2018 Edition of IEEE 1584 was to have a model that can provide reproducible results and greatly improved accuracy.

Reality: Keep in mind the tests and equations are based on an empty boxes with electrodes of specific gaps and orientations. That means the results may vary depending on the actual equipment. Also keep in mind, there can be multiple gaps and electrode configurations for any give piece of equipment.

Usual disclaimer: This is not an official representation of IEEE 1584. Just a personal comments from me.
Author:  bbaumer [ Mon Feb 13, 2023 1:39 pm ]
Post subject:  Re: How accurate is arc calculation really
Jim Phillips (brainfiller) wrote:
I did not sift though all of the data but recall one of the many objectives for the 2018 Edition of IEEE 1584 was to have a model that can provide reproducible results and greatly improved accuracy.

Reality: Keep in mind the tests and equations are based on an empty boxes with electrodes of specific gaps and orientations. That means the results may vary depending on the actual equipment. Also keep in mind, there can be multiple gaps and electrode configurations for any give piece of equipment.

Usual disclaimer: This is not an official representation of IEEE 1584. Just a personal comments from me.


No juicy stories about the time you almost shut down Wakanda during a test that didn't go as planned?
Author:  Jim Phillips (brainfiller) [ Wed Feb 15, 2023 7:45 am ]
Post subject:  Re: How accurate is arc calculation really
bbaumer wrote:
Jim Phillips (brainfiller) wrote:
I did not sift though all of the data but recall one of the many objectives for the 2018 Edition of IEEE 1584 was to have a model that can provide reproducible results and greatly improved accuracy.

Reality: Keep in mind the tests and equations are based on an empty boxes with electrodes of specific gaps and orientations. That means the results may vary depending on the actual equipment. Also keep in mind, there can be multiple gaps and electrode configurations for any give piece of equipment.

Usual disclaimer: This is not an official representation of IEEE 1584. Just a personal comments from me.


No juicy stories about the time you almost shut down Wakanda during a test that didn't go as planned?


Hum, stories? Yes, I have a few arc flash tests go down the path of the unexpected. Here are two.

Burn down the lab
This was the best one ever for me. I was brought in by a research team for a Japanese delegation to determine how a nuclear plant caught on fire after the Fukushima earthquake of 2011. At the initial meeting with the Japanese delegation, I suggested a 9.0 earthquake might shake the switchgear so violently that the bus could make momentary contact and initiate an arc flash. They had not heard of an arc flash at that time so a few demo videos later and we were on our way for a recreation later that year.

One of the tests gave them exactly what they wanted. An arc flash catching the MV switchgear on fire – aftermath in the photo. There was a wireway at the top packed with the type of conductors used at the plant. The 3 second arc flash ignited the conductor insulation and the wireway acted like a chimney with a perfect draft going.

With flames shooting towards the ceiling (see photo) the delegation asked me how to put the fire out. With a very straight face I stated, I have no idea. I was brought in to blow it up, not extinguish it. After a panicked 15 minutes, someone with a garden hose and lift extinguished the fire by flooding the wireway. The damage looked quite similar to the photos of the switchgear from the nuclear plant. A success! (with some excitement)

Attachment:
Test Fire.png
Test Fire.png [ 2.35 MiB | Viewed 16171 times ]

Attachment:
SWGR EDIT.png
SWGR EDIT.png [ 34.4 MiB | Viewed 16171 times ]


MCC 2 for 1
Another was an MCC section. The arc flash was staged at the top of the vertical bus where the incoming 480V supply was located. High speed video shows the arc running towards the bottom – away from the source - as expected. Except, it began at the top once again and repeated the run to the bottom. A “2 for 1” arc flash. What?? All we could assume is some of the plasma made its way towards the top to reignite and begin it all over again. A bit of a head scratcher.

This is one reason it is difficult to provide a black and white answer regarding arc flash. On some occasions, there can still be still some unpredictability with it.
Author:  bbaumer [ Wed Feb 15, 2023 8:00 am ]
Post subject:  Re: How accurate is arc calculation really
Jim Phillips (brainfiller) wrote:
bbaumer wrote:
Jim Phillips (brainfiller) wrote:
I did not sift though all of the data but recall one of the many objectives for the 2018 Edition of IEEE 1584 was to have a model that can provide reproducible results and greatly improved accuracy.

Reality: Keep in mind the tests and equations are based on an empty boxes with electrodes of specific gaps and orientations. That means the results may vary depending on the actual equipment. Also keep in mind, there can be multiple gaps and electrode configurations for any give piece of equipment.

Usual disclaimer: This is not an official representation of IEEE 1584. Just a personal comments from me.


No juicy stories about the time you almost shut down Wakanda during a test that didn't go as planned?


Hum, stories? Yes, I have a few arc flash tests go down the path of the unexpected. Here are two.

Burn down the lab
This was the best one ever for me. I was brought in by a research team for a Japanese delegation to determine how a nuclear plant caught on fire after the Fukushima earthquake of 2011. At the initial meeting with the Japanese delegation, I suggested a 9.0 earthquake might shake the switchgear so violently that the bus could make temporary contact and initiate an arc flash. They had not heard of an arc flash at that time so a few demo videos later and we were on our way for a recreation later that year.

One of the tests gave them exactly what they wanted. An arc flash catching the MV switchgear on fire – aftermath in the photo. There was a wireway at the top packed with the type of conductors used at the plant. The 3 second arc flash ignited the conductor insulation and the wireway acted like a chimney with a perfect draft going.

With flames shooting towards the ceiling (see photo) the delegation asked me how to put the fire out. With a very straight face I stated, I have no idea. I was brought in to blow it up, not extinguish it. After a panicked 15 minutes, someone with a garden hose and lift extinguished the fire by flooding the wireway. The damage looked quite similar to the photos of the switchgear from the nuclear plant. A success! (with some excitement)

Attachment:
Test Fire.png

Attachment:
SWGR EDIT.png


MCC 2 for 1
The first was an MCC section. The arc flash was staged at the top of the vertical bus where the incoming 480V supply was located. High speed video shows the arc running towards the bottom – away from the source - as expected. Except, it began at the top once again and repeated the run to the bottom. A “2 for 1” arc flash. What?? All we could assume is some of the plasma made its way towards the top to reignite and begin it all over again. A bit of a head scratcher.

This is one reason it is difficult to provide a black and white answer regarding arc flash. On some occasions, there can still be still some unpredictability with it.


Those are great. Thanks for sharing. Much appreciated.
Author:  Felix Nepveux [ Tue Feb 21, 2023 9:32 am ]
Post subject:  Re: How accurate is arc calculation really
The calculation is exact, the data is the problem. The data is actually not truly "exact", it is a curve fit from thousands of tests, not a scientifically exact calculation, but the programs will give the same results for the same data. The world changes, and there will be different arcs and energy releases from the same location, the "Calculation" procedures are just the best we have to try to insure safety on a consistent basis. It is like standard of care in medicine, once bloodletting was the standard.
Author:  bbaumer [ Tue Feb 21, 2023 11:10 am ]
Post subject:  Re: How accurate is arc calculation really
Felix Nepveux wrote:
The calculation is exact, the data is the problem. .


I think this is an accurate statement as-is Jim's observation about "real world". In the control panel example there were dozens of combinations of bus gaps and buss configurations and several different conductor sizes from 3/0 to #12. Multiple breakers, relays, MCP's, starters and even a VFD inside it (it was a large control panel). How do you really accurately calculate something like that? The steel conduit was the equipment grounding conductor. Was every joint made up tightly and corrosion-free? What if it wasn't? What about all the paths through the steel building structure? How do you calculate that if the conduit path itself isn't exactly low impedance?

Somewhere on this forum from a few years ago somebody posted some test data from EPRI when they shorted out the secondary side of pad mounted transformers and could not get an arc to sustain itself because the bus gaps between the paddles/spades were too large. Does that mean there is no arc flash hazard there instead of what we often calculate as a very high incident energy at that point? Probably not.
Author:  Doug Powell [ Tue Mar 21, 2023 9:41 am ]
Post subject:  Re: How accurate is arc calculation really
It seems to me that empirical studies are very useful for two reasons, first to formulate a hypothesis, and second to validate accuracy of a model. The first inherently allows for some variation and the second is inherently problematic. Someone, somewhere can always come up with a special case. Those "one off" anecdotal cases are statistically rare but make for great war stories. So I would say the final answer in system design comes down to question of cost vs benefit vs risk.

Bear with me. My principle occupation is as an electrical safety engineer in large scale energy storage systems, often with batteries scaled up to multiple megawatt/hours. When I talk to people about this topic, I sometimes use the analogy of the wooden kitchen match you strike on the box. This match provides a certain utility people want or need along with a level of hazard which is expected, going back to the early 1800s. An untrained user or child may inadvertently drop a recently used match in the trash can, with a burning ember. While it is possible to design a match to self-extinguish within a predetermined number of seconds, it would be extremely expensive and no one would buy it.

So my question is, how much safety margin or variance in the model do we account for without exceeding some perceived limit by the designer, end user or a court of law?

~Doug
Page 1 of 1 All times are UTC - 7 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/